<b>General descriptive studies (00350)</b><br>We have used incidence and mortality data from the Surveillance, Epidemiology, and End Results (SEER) program to investigate demographic patterns. An analysis of SEER data for 26,758 cases of soft tissue sarcomas regardless of primary site diagnosed during 1978-2001 found that almost half (47.9%) arose in the soft tissues; rates varied markedly by race, gender, age, and histologic type, suggesting that these tumors may be etiologically distinct. The new WHO classification scheme considers B-cell chronic lymphocytic leukemia (CLL) and small lymphocytic lymphoma (SLL) in an aggregate category (CLL/SLL); however, our analysis of the SEER data revealed similarities and dissimilarities in the incidence patterns for the two malignancies. Based on registry data from 41 populations in 14 countries, cohort-specific trends in testicular seminoma and non-seminoma appeared similar, suggesting that subtypes are epidemiologically and etiologically comparable. Additional analyses are assessing the incidence patterns according to anatomic subsite and histopathologic type for several cancers including esophageal cancer, stomach cancer, lung cancer, and kidney cancer.<br><br><b>Special descriptive studies (10348)</b><br>Epidemiology studies have generally viewed breast cancer as a single biologic entity with common etiology and unified pathogenesis. Accumulating data challenge this view, suggesting that breast tumors may be categorized into several groups with distinctive epidemiological features, clinical characteristics, and outcomes. Comparison of female breast carcinoma incidence rates among native Japanese in Osaka, Japanese-Americans, Whites and Blacks in the U.S. revealed age-specific incidence differences among Occidental and Asian breast cancer populations. Age-specific rates among women in the U.S. reflected bimodal early-onset and late-onset breast cancer populations, whereas rates in Japan had mostly early-onset age distributions-at-diagnosis. Using data from a population-based case-control study in Poland, results confirmed etiologic heterogeneity by age-at-onset for certain risk factors such as parity. Parity was protective for late-onset breast cancers, but a risk factor for early-onset tumors. The reversal of relative risks by age at onset is a qualitative (crossover) age interaction, suggesting that breast cancers are fundamentally divisible into at least two main types. The 1<sup>st</sup> breast cancer is early-onset and influenced by etiologic events, occurring early in reproduction life. The 2<sup>nd</sup> breast cancer is late-onset and impacted by life long carcinogenic exposures.<br><br><b>SEER special studies (00316)</b><br>Analysis of gene expression profiles in breast cancer have identified intrinsic molecular subtypes, which differ in risk factor profiles and prognosis. These advances in molecular taxonomy, prevention, and treatment create a need for establishing the incidence and prognosis of specific breast cancer subtypes in various populations. However, until recently, there were limited population-based resources for these estimates. In 2001, the SEER program supplemented tumor registries to collect discarded formalin-fixed, paraffin-embedded tissue blocks from pathologic laboratories within their catchment areas. In a demonstration project, we validated the utility of SEERs Residual Tissue Repository for molecular markers, using an existing set of breast cancer tissue microarrays (TMAs).<br><br><b>Mortality Rate Generator Software (00390)</b><br>The online version of the Atlas of Cancer Mortality in the United States, 1950-94, published in 1999, is available at http://www.nci.nih.gov/atlasplus. Users can create customized maps according to cancer, age groups, sex, and race.<br><br><b>Evaluation of Disparities in District of Columbia (DC) (10301)</b>Newly-emerging data from the population-based DC Cancer Registry provide a unique opportunity to investigate disparities in incidence rates according to demographic, geographic, and socioeconomic characteristics. Total cancer incidence was higher among blacks than whites by 53% among men and 2% among women.<br><br><b>Record Linkage StudiesSweden and Denmark linked registries on hospital discharges and subsequent cancers (00560)</b><br>A Danish case-control study of intrahepatic cholangiocarcinoma (IC) and population controls selected from the Danish Population Register were linked to the DHDR to obtain exposure/hospital data on prior hospitalizations. IC was significantly associated with alcoholic liver disease, chronic inflammatory bowel disease, and bile duct diseases.<br><br><b>Swedish CER occupational study (02050)</b><br>Using Swedish census data from 1960 and 1970, linked to the Swedish Cancer Registry for cancer ascertainment from Jan 1, 1971 to Dec 31, 1989, occupational physical activity was found to be associated with a decreased risk of colon cancer among Swedish men and women, particularly the proximal and middle parts of the colon among women and distal colon among men.<br><br><b>Swedish childhood cancers study (00550)</b><br>A study of childhood bone cancers in the Swedish birth registry dataset found several risk factors implicating complicated delivery and fetal distress. We are also studying perinatal risk factors for neuroblastoma.<br><br><b>Veterans Administration hospitalization database, Patient Treatment File, and Outpatient Clinic File (00580)</b><br>A cohort of 1,500 patients with Gaucher disease was found to have 2-3-fold increased risks of non-Hodgkin lymphoma, malignant melanoma and pancreas cancer, but no significant association with multiple myeloma or cancer overall. An analysis of medical risk factors for chronic lymphocytic leukemia (CLL) found increased risk associated with chronic sinusitis, pneumonia, Herpes zoster and simplex, auto-immune hemolytic anemia and prostatitis.<br><br><b>US Military Cancer Institute (USMCI)/NCI Collaborative Research Program (10382)</b><br>DCEG and USMCI researchers are analyzing data on more than 9 million active and retired military personnel and their families to estimate cancer rates as well as study the effects of occupational exposures and lifestyle factors on cancer risk. The wealth of information, including more than 30 million serum samples, will allow valuable studies, including research on rare cancers, cancers common among younger men, and cancers that occur more frequently in minority populations